With sweeping changes in customer expectations happening every now and then, it is only fitting for innovative companies to predict these demands ahead of time. A global leader in HVAC(Heating, ventilation and Air conditioning) and home appliances wanted to execute its renewed focus on innovation, the company wanted to stay ahead of its supply chain requirements and market expectations.
-
The Objective: Building a Next-Gen Forecast Workbench
- Accelerate demand forecasting through a forecast workbench
- Enhance the accuracy of forecasting with AI and ML algorithms
- Automate the existing S&OP processes through an intuitive and customizable dashboard
- Improve visibility across the supply chain for all relevant stakeholders – both internal and external
Key Challenges
- For our client, conventional and manual demand forecasting methods no longer sufficed, given its environment of working with long supply chains, short product lifecycles, and volatile demand. Faced with these challenges
- Company wanted to improve the accuracy of its demand forecasting to ensure its suppliers, logistics providers, and, more importantly, products and customers are on the same page.
- They also wanted to improve their sales and operations (S&OP) processes that were predominantly spreadsheet-based and lacked comprehensive visibility and timely accuracy. It was a challenge for demand planners and other stakeholders involved in the process to predict demand and amend changes on trends in real time.
Ready to experience?
TALK TO EXPERTS
The Solution
AI-Generated Forecasts Leveraging Amazon Forecast and Visualizing it in a Workbench
Infosys partnered with the HVAC enterprise to build the demand forecasting workbench on Amazon Forecast, and used AI and ML to predict demand consistently and accurately for a longer forecast horizon of 18 months. The monthly forecasts generated are supported by an AWS Glue-hosted Python data processing pipeline with data storage in Snowflake and Amazon S3. Some of the striking highlights of the demand forecasting workbench are:
- Display historical actuals and 18-month future forecasts
- Adjust forecasts during S&OP meetings in a three-stage workflow
- View aggregated forecasts at a category and plant level, quarterly or annually
- Export forecasts to spreadsheets
Benefits
Infosys helped the client to:
Reduce the demand forecasting timeline from days to few seconds
Enable reallocation of valuable human power to more pressing needs of extracting insights from the forecasts
Improve user experience during S&OP discussions
Enable substantial revenue increase and cost savings
